Skip to main content

Full text of "Notes on making saltpetre [sic] from the earth of the caves"

See other formats

Notes on making 
Conf Pam l2mo #417 







Bt major GEO. W. RAINS, 


In charge of the Gunpowder Department, C. S. A., late of the U. S. A., and former Asst. Prof. 
of Chemistry, &c., U. S, M. A. 



The crude saltpetre from the caves — called grongh salt'petre in com- 
merce — requires to be purified before it can be used for gunpowder, and 
for this purpose government has established a refinery at Nashville 
capable of refining daily 5000 pounds of grough saltpetre into pure 
nitre, as white as snow% and ready for the pow^der mills. In the exten- 
sive Government Powder Works, now in course of rapid erection in 
Georgia under the direction of the writer, over five tons of saltpetre v/iil 
be refiued each day, if required, and converted into gunpowder. 


One ordinary iron pot, for boiling ; three or four tubs, pails, or 
barrels cut off ; two or thi-ee small troughs ; some coarse bags or a 
wheelbarrow to bring the earth from the cave, and four strong barrels 
with one head in each — empty vinegar, whiskey or pork Ijarrels are very 
good— are about all the articles required for a small saltpetre manufac- 
tory. To these, however, must be added some ash barrels to make 
potash lye, as it is better that this should be made at the same time and 
place, the ashes from the fire under the pot for boiling assisting in the 
production . 


First bore a hole about the size of the finger through the head or end 
of each barrel near one side, and fit a wood plug to each hole — then set 
the barrels on some pieces of timber near each other, the heads down, 
and the hole of each projecting over the timber. Put some twigs into 
the bottom of each barrel, and on these place straw or hay about half a 
foot thick when pressed down ; then, having brought some of the earth 
from the cave, and broken up all the lumps, fill each barrel full without, 
pressing it down. Put the plugs into the holes tightly, and fill up each 
barrel with as much water (hot water is best m winter j as it will hold ; 
allow the whole to remain until next day, then pull out the i)lugs, hav- 
ing: placed a tub or pail under each, and pour all the water from the 
first barrel into the second barrel, and all the water or liquor v/hich 
drains from this 1)arrel must l)e poured on top of the earth of the third 
barrel, and finally the liquor which drains from this last barrel must be 
poured into a tub or other vessel. Now having previously made some 
strong lye from wood ashes, pour a small stream of it into the tub and 
scir it well ; immediately the clear liquor will become muddy, and as 
long as the lye contiimes to curdle or cloud the liquor, it must be poured 
in ; of course you will have to wait now and then for the licjuor to settle 
to see if it requires more lye. No more must be used than is necessary, 
for it not only v-astes the lye, but is an impurity which the refinery must 
afterwards get rid of. We will suppose that the proper quantity of lye 
hns been used, aud the liquor allowed to settle or drain through cloth 
until it becouaes clear ; it is then poured into the pot and boiled away 

until a drop taken up by the end of a stick becomes hard or solid when 
let fall upon coldraetal or upon a plate. 

The liquor is now to be dipped out of the pot and poured into a cloth 
placed over a tub or barrel, and allowed to strain through into the tub 
below and become cold. As soon as the liquor begins to cool, crystals of 
snltpctre will commence forming, and when cold the liquor left — called 
mother liquor — must be poured off from the saltpetre back into the pot 
with the fresh liquor for Iwiling, as it still has considerable saltpetre in 
it. There will be found at the bottom of the pot after the liquor is 
dipped out, when the boiling is completed, some earthy salts which, after 
draining, can be thrown away as impurities ; if, however, some long 
needle-shaped crystals should be seen in it when cold, it contains some 
saltpetre, and about a quart of hot water should be added, and then 
poured off after a time, when it will have dissolved all the saltpetre 
left among the earthy salts ; this wash water can then be put back into 
the pot after the impurities shall have been cleaned out. 

The Saltpetre formed by the foregoing process must be first allowed 
to drain well, and then placed on cloths stretched before the fire or out 
in the sun to dry ; when the drying is completed, it is to be put into 
sacks or barrels, and is ready to be transported to the Government Agent 
at Nashville, Lieut. M. H. Wright, C. S. A. ordnance officer, who will pay 
for the same on receiving the bills of its shipment on the railroad. 

If the crystals of saltpetre are wet and brown, and will not keep dry, 
it is because too much lye from the wood ashes has been used ; this can 
be removed by nearly filliug a tub or barrel with the saltpetre and 
pouriug cold water on it, as much as the tub will hold, and after remain- 
ing about one hour, the water can be drained off from the bottom, when 
it will carry with it most of the lye ; this wash water must be poured 
iuto the lye of the wood ashes so as not to lose the saltpetre which it 

The foregoing process evidently contains all that is required in prin- 
ciple for the making of Saltpetre on a large scale, since nothing more is 
to be done than to increase the number of barrels and boilers. Casks 
would be better perhaps than barrels in such case, and vats made by 
placing the lower ends of pieces of plank about four feet long into a 
trough, and opening or spreading out the upper ends about three feet, 
then making ends to the vat, is an economical and convenient arraage- 
ment, which may be used on a small as well as a large scale, instead of 
casks or barrels. In making use of these vats, strips of wood should be 
placed over the edges of the planks on the inside, and a thick layer of 
twigs and straw should be placed at the bottom between the planks on 
the inside, as well as along the sides and ends to about one foot of the 
top of the vat, and kept there by pieces of wood leaning against the 
sides, whilst it is filled with earth from the cave ; if this is not done, the 
liquor will in many cases drain through very slowly, and time lost to no 
purpose. A hollow cr channel about a foot deep should be made along 
the centre of the earth in the vat, to collect the water poured in. 

Whether vats, casks or barrels be uied, the same principle must be 


carried ont of passiug the leached (or drained) h'qnor from the first ves- 
sel into the second, and from the second to the third before boilmg, 
otherwise there will be much time and fuel lost in.neleas boding of a 
weak liquor ; this is a common error at the caves, and causes the salt- 
petre to cost more than is necessary in time, labor and fuel. 

We will now follow the process of leaching more particularly. Sup- 
pose it takes eight gallons of water to fill up the barrel after the earth 
has been put in even with its top, or nearly so, then about oue-half or 
four gallons onlv, will drain off, generally ; we must now refill the bar- 
rel with four gallons more of water, and this time four gallons will dram 
or leach out, because the earth has already been charged with water. 
Again we refill the barrel the third time, putting in four gallons of water 
more, and after four gallons of liquor (or as much as will drain away) 
has again leached off,"the earth must !)e thrown out, and the barrel re- 
filled with fresh earth from the cave. 

It .will now be explained how to proceed so as to have a regu- 
lar rotation of the barrels, as they shall be emptied one after 
the other of the old earth and refilled with fresh earth from the 
cave. It will be supposed at first that the work is on a small scale, 
then there • will be required four barrels, if it be desired to proceed 
economically; to make it clear, we will suppose that these four bar- 
rels are placed round in a circle near each other, and three of them 
are filled with fresh earth, the fourth remaining empty. Now when 
the earih of the first barrel has been exhausted of saltpetre by the three 
washings, it will be thrown out, but instead of filling up this barrel 
with frcsii earth, we fill up the fourth or empty barrel, and this can be 
going on during the leaching.- Thus we have three barrels working as 
at first, the fresh barrel being the third in the new series, and receiving 
the leached liquor from the one iK-xt to it. The fir?^t barrel of the new 
arrangement, however, has already been washed twice, before the new 
barrel was filled with earth ; hence, after washing it once more with the 
four gallons of water, which it has just received from the barrel just 
emptied^ — which quantity drains off and is poured into the one next to 
it— the earth is removed and this barrel left empty. Barrel number 
one, which we first, emptied of the old earth, having now been refilled 
with earth from the cave, becomes the third in the new arrangement, 
and so on. 

Tne liquor of the first of the three working barrels being always 
poured into the second barrel, and the liquor which leaches from this 
to be always poured into the third barrel, and finally the liquor which 
drains from this barrel is to be put into the vessel, where the lye from the 
ash barrel is mixed with it and the whole allowed to settle. When the 
liquor has become clear by settling or being strained through a cloth, it 
is ready to be placed into^he kettle for boiling down. Thus there is a 
continual rotation of the work of the barrels without disturbing anything 
and r-onstantly providing strong liquor for the ketth^ to be boiled with- 
out losing any saltpetre in the earth thrown away. Ft will generally 
take two'^days for the liquor tO drain off from each barrel, but the timo 

will vary with the nature of the earth, as a saudy one may take but a few 
hours, whilst a clayey one may take tliree or more days. In this case it 
would be better to mix sand with the earth, leached ashes or gravel, or 
even hay or straw, than to lose so much time. If vats be made deeper 
than a barrel, more time will be required to leach them oflf, which, oi 
course, is to he avoided, as nothing is gained by taking two or three 
weeks to leach off a large vat, whilst the same earth in smaller vats or 
casks may be leached off in two days. 

With I'egard to mixinsr the lye of wood ashes with the liquor of the 
third vat or barrel, the proper way is to take a certain quantity, say one 
pint of the liquor, and put it in a clear glass, then gradually add the lye 
and stir well. So Ioulc as the lye curdles or clouds the liquor more must 
be added. When sufficient lye has been used, allow ihe liquor to settle 
and become clear, then add a few drops more of lye, if it no longer 
clouds the liquor, sufficient has been used, and if adding the lye has been 
done carefully, no more has been employed than was just necessary to 
precipitate the impurities. We will suppose one gill of lye has been 
used to the pint of liquor, then it would take eight gills or one quart of 
lye to eight pints or one gallon of liquor — hence, knowing the number 
of gallons of liquor, it is easy to see at once how many gallons of lye 
must be added without further trial 

The above experiment may he employed to ascertain if any earth 
contains Saltpetre, for if the lye of wood ashes causes a curdling, or 
muddies the water in which a considerable quantity of the earth has 
been mixed, and then drained olF, we may presume that there is Salt- 
petre present, and the quantity will, in general, bt- in proportion to the 
amount of curdling. A slight clouding of the liquor may be produced 
by other salts Ix'ing present instead of Saltpetre, but if there is much 
curdling, it is a pretty sure sign of Saltpetre. If there be any doubt, 
however, allow the muddy liquor to settle, then draw it oft* and boil down 
until it thickens, then dip a slip of paper into it and dry it well, touch a 
coal of fire to the paper, and if it burns rapidly and sparkles, you may 
be sure Saltpetre is present. 

In making lye from wood ashes it is well to remark that the leaves, 
bark, branches and liml)s of the tree contain more potash than the trunk 
and that the oak and ash are generally the best woods to get ashes from, 
in leaching the ashes a similar arrangement may be used, as for the 
Saltpetre liquor, that is, four Ijarrels may be used together in the same 
way, keeping three filled with ashes and the fonrth empty, and passing 
the lye from one to the other as before explained. By this means 
stfong lye is always on hand to be used, and the Saltpetre liquor is not 
watered too much by a weak lye which, lias to be evaporated away at an 
expense of fuel and labor. 

I will now speak of the economy of labor in the operations : thus, if 
the cave is sufficiently large and hght enough, or can be lighted cheaply 
by fire — the ashes of which may be used — it is evidently cheaper to 
carry the vats and Ijoilers into the cave just where the earth is, or near 
by, than to carry the earth out in. bags to the outside of the cave to be 


leached. Again, it is more economical to bring water to the earth than 
to jtake the earth to the water, when both are outside of the cave, as is 
sometimes done, because there is much less weight of water used. By 
pursuing the method pointed out no time or material is lost — each day 
has its regular recurring operations to be performed — no surplus water 
has to be boiled away, and no lye is wasted, rendering the Saltpetre 

To give some idea as to the quantity of Saltpetre that can be made, 
I will state that twelve barrels of the earth of the caves will, in general, 
make not less than one hundred pounds of Saltpetre, and this will take 
from twelve to fifteen bushels of ashes. 

If the twelve barrels are arranged in four circles or rows, with an extra 
or fourth barrel to each row, then a barrel of the leached earth can be 
emptied from every other row each day, (or two barrels a day) and the 
same number filled with fresh earth, thus in six days the twelve barrels 
will have been worked through, and this can be done by one man, whilst 
a second man boils away the liquor and attends to the vats or barrels ; a 
third man can more than supply the ashes used, and can assist in filling 
the barrels. Thus in six days we have the labor of three men, which is 
abundantly ample to make over one hundred pounds of Saltpetre, unless 
the earth is difficult to be procured. If the earth is richer than that 
supposed, or if the leaching of each barrel takes but one day, instead of 
two, as will frequently be the case, then two hundred pounds of Saltpe- 
tre may be procured in the same time, but in the latter case will require 
additional labor. This is at present worth seventy dollars, being the 
price which Government now pays for a limited time, at thirty-five cents 
per pound, to encourage its production and to remunerate individuals 
for first cost of apparatus. 

Those who manufacture Saltpetre on a considerable scale will find it 
convenient to have two or more casks or cisterns sunk in the earth, to 
receive the mother liquor from the evaporating kettle, where it is left 
for twenty-four hours to crystalyze its Saltpetre. 

In boiling the liquor from the vats or barrels, after it has settled or 
been strained from the sediment formed by adding the lye, a thick skum 
will rise to the surface, which must be skimmed off, as it forms, and 
thrown on the top of one of the vats, so as not to lose 'the Salpetre that 
may be dissolved from it. 

The bottom of the pot or boiler, after a time, will become foul from 
the lime and earthy salts deposited on it, which can be prevented, if 
thought necessary, to a considerable degree, by sinking in the kettle a 
small pot with a wide mouth. The sediment will collect in this pot 
and can be removed from time to time, because the liquor remains com- 
paratively still within it and allows the salts to settle, whilst the agitation 
of the boiling prevents the sediment falling io the bottom of the kettle. 

Saltpetre made after the foregoing directions will not have above five 
per cent, of impurities ; but if carelessly made it will have much more, 
and as these have to be separated at the refinery, before it can be used 
for gunpowder, such Saltpetre is not worth so much to Government. 


Tasting the earth to see if it will yield Saltpetre is not a very accu- 
rate way of determining the fact, because the lime Saltpetre has less 
taste than the Potash Saltpetre, and the former is the one mainly in the 
earth, which lye converts into ordinary, or potash Saltpetre. Thi:s 
an individual might be deceived into rejecting earth which may yield 
a snfficient quantity, if worked. 


In order U> call attention to the very consideral)le loss sustained by 
imperfect working of the earth of the caves, I will state that at a certain 
cave in Georgia, which was examined by Professor Pratt, of the Ogle- 
thorpe University, who kindly furnished me with the result of his 
analysis — it appeared that in earth which actually contained not less, on 
the average than 90 pounds to the barrel — nuich of it containing: 120 
jjounds — only about 67 pounds were obtained. Thus one-fourths at least 
of the entire amount of Saltpetre was lost, or about eight dollars to each 
barrel ; also, the amount of labor employed was abundantly ample to 
have obtaiufHl and leached, daily, twice the quantity of earth that was 
done, of which there was sufficient within 200 yards of the mouth of the 
cave to furnish forty-five thousand pounds of Saltpetre. Hence the loss 
at this cave was as follows, for each 12 barrels of earth actually leached 
at this time. 

'i'welve barrels of earth or 90 pounds of Saltpetre lost to one 

barrel leached, which, at 35 cents, amounts to - - $31 50 

One quarter of the Saltpetre lost to each leached barrel, or 22^ 

pounds, at 35 cents, amounts to 7 87^ 

Total lost, - $39 37i 

Here was an actual loss to the proprietor of nearly forty dol- 
lars ; thus be received for each 12 barrels worked, obtaining 
67 pounds only, $23 45 

Should have received, with proper working, with the same 
amount of labor and capital, in tlu; same time 24 barrels, or 
180 pounds, $63 00 

The above case is a sufficient demonstration of the necessity of pursu- 
ing the method laid down in these notes for the making of Saltpetre. 




Two evaporating kettle or sugar pans capable of containing about 
forty gallons each ; one kettle or boiler holding not less than twenty- 
five gallons ; one barrel arranged with a hole and ping at bottom, and 
covered loosely with two thicknesses of bagging, or coarse cloth, at its 
open end, forming a bag for straining ; one shallow wooden trough six 
feet long, three feet broad, nine inches deep, for cooling ; one wooden 
rake ; one spade or shovel, having a. long handle ; one wooden straining 
box or trough, three feet three inches long, twenty inches broad and six 
inches deep, with several small holes in its bottom — this box is placed on 
the top of the long trough, at one end ; one wash barrel having a second 
bottom pierced with holes about three inches above the true bottom, this 
second bottom is to be covered with coarse cloth — hetween the bottoms 
a hole and plug are made ; one cask to receive wash water ; one cask or 
barrel nearly filled with water to receive all the refuse Saltpetre, and in 
which the old filtering cloths are th'own to dissolve out the Saltpetre ; 
one cask or large barrel to receive mother liquor ; one platform scale or 
set of steelyards ; together with some buckets, drying cloths, &c 


Weigh out two hundred and twenty-five pountls of Saltpetre and put 
it into the kettle or boiler, with sixteen gallons of water ; light a fire 
under the kettle and let it boil — not too briskly, however— for about two 
and a half hours, removing the skum which rises to the surface, which 
should be thrown into an e'mpty barrel. Cold water must be thrown in 
occasionally to keep the liquor to the same height in the kettle, for it 
must not be allowed to boil away. After the boiling is finished, allow 
the fire to die out, and dip the liquor — not allowing it to cool — into 
the cloth on the top of the straining barrel, whence it is allowed to run 
into the long cooling trough ; here it is constantly agitated by raking it 
forwards and backwards by means of the wooden rake, until it has cooled 
down to about blood heat, which will take probably two hours oi- more. 
During the time of cooling, large quantities of fine needle-shaped crys- 
tals of nitre will form in the liquor, which are to be taken out by means 
of the long handled spade, and thrown into the d'-aining trough on the 
end of the cooling trough. When the liquor has sufficiently cooled 
down, run it off into a cask sunk into the earth for tlaat purpose, by 
means of a hole and plug in one of the lower ends of the cooling trongh 


The crystals of nitre in the draining trough will now commence look- 
ing white as snow, and are to be left to drain until next day, when the 
nitre is removed to the washing barrel, which should be cut off at such a 
height as shall be about half filled with the crystals. 

This barrel is then to be gently filled with cdd water to the top, and 
allowed to remain one hour, when the plug is taken out, and the liquor, 
which is nearly saturated with nitre — holding in solution all that re- 
mained of the mother licjuor — is allowed to drain off into the cask kept 
for that purpose. The nitre thus made is nearly pure, sufficiently so for 
nearly all purposes, and can be made into gunpowder. To make the 
finest quality of powder, however, the crystals must l)e twice washed 
before being taken from the washing barrel, cold water being poured in 
each time until the barrel is full, and after remaining one hour each time, 
is to be drawn off as defore, and the nitre well drained and then dried ; 
the crystals are now entirely pure, and can be used for the best quality 
of gunpowder. 

The foregoing is the process, on a much larger scale, pursued at the 
Government Refinery, under direction of the writer, and is a great im- 
provement on the old process, taking only one-sixth part of the time for- 
merly consumed, and hence saving largely in time, labor and fuel. It 
is, in the main, the method pursued at the celebrated Government 
Powder Works at Waltham Abbey, England. The writer is now en- 
gaged in making some experiments by which he anticipates the process 
will be considerably shortened, thus enabling the Government Refinery 
to double its daily product without increasing the apparatus. 

It must be observed that in recharging the boiler with Saltpetre, in 
stead of putting in the previous amount of 225 pounds, only 200 pounds 
will be used, because in place of pouring in fresh water, as in the first 
case, we will now make use of 16 gallons of the wash water from the 
crystals, which holds about 25 pounds of nitre in solution. 

This is under the supposition that the temperature of the wash water 
is about 65 degrees ; but if it is colder than this, it will contain less 
nitre, and if the temperature be that of freezing, only about nine pounds 
of nitre will be found in the 16 gallons, instead of 25 pounds. 

Where two washings take place, I find it much more economical in 
fuel to have a separate cistern to hold the liquor of the second washino-^ 
which is nearly a pure saturated solution of nitre, and this is used for the 
first washing in the next process ; thus is saved the evaporation of a 
large quantity of water, which would require additional evaporating 
pans and furnaces. I find no appreciable difference in the purity of the 
nitre thus washed from that produced by the mode of washing "of Wal- 
tham Abbey. 

That portion of the wash water which is not used in the boiler with 
the new charge of Saltpetre, is to be removed to the evaporating pans 
with the mother liquor. In the cask containing the mother liquor from 
the cooling trough, there will be found next day a considerable amount 
of large crystals of saltpetre, which can be collected and thrown in with 
the grough Saltpetre. The method of evaporating the mother Uquor, 


and crystalizing its saltpetre, is entirely analogous to that already ex- 
plained in making saltpetre. 

It may be observed that about a gallon of liquor should be taken out 
each day from the waste cask and put into the evaporating kettle, whilst 
the same amount of fresh water should be poured iuto the cask, in order 
to prevent the water in this cask becoming saturated with the waste salt- 
petre which is from time to time thown into it, as also what it acquires 
from the soaking of the filtering or straining cloths, etc. When the bar- 
rels into which the skimmings are thrown becomes full, it is to be poured 
into a cloth placed over the cask containing the mother liquor:; being 
drained, a small quantity of hot water should be poured over what re- 
mains, and then the refuse may be thrown away. 

If a larger or smaller quantity of saltpetre be refined than that men- 
tioned, then corresponding proportions of the saltpetre and water will be 
employed ; thus in the Government Refinery at Nashville, 5000 pounds 
of the salt are used with generally about 360 gallons of water, which are 
boiled together for four hours. The amount of impurities shonld regu- 
late the amount of water used, but this is not of much moment in small 


Refined Saltpetre is not required from the caves that is done by Gov- 

It was stated in the body of these notes that Saltpetre should be 
put up in bags or barrels for transportation ; it may also be put up in 
kegs or strong boxes, the latter being made about two feet long and 
fifteen inches square (section) well nailed. 

The Saltpetre may be sent to any ordnance agent that may be conve- 
nient, as will be seen by the advertisement below. 


The Ordnance Department, Confederate States, will pay thirty-five 
cents per pound for all Saltpetre delivered be fore the first of February, 
1862, at any of the following points ; 

Capt. W. G. Gill, Augusta, Ga. ; C. G. Wagner, Military Store 
Keeper, Montgomery, Ala. ; Lieut. M. H. Wright, Nashville Tenn. ; 
Capt. W. R. Hart, Memphis, Tenn. ; Sandford C. Faulkner, Military 
Store Keeper, Little Rock, Ark., and at Richmond, Ya. 

J. GORGAS, Lieut. Colonel, 

Chief of Ordnance. 

HoUinger Corp.